Author Topic: ($40 GPSDO working well now, my mistake) (Read 61945 times)

Miti · « **Reply #225 on:** December 13, 2018, 05:36:59 pm »

Quote from: cdev on December 05, 2018, 04:29:57 am

Our goal should be making it possible for everybody to have a decent standard of living and security on as little money as possible. Not make poor people die earlier from stress.

How many lives does it save to make taxpayer developed cancer drugs that don't actually improve on existing ones at all, more expensive by evergreening patents and eliminating alternatives? None. It kills poor people. It steals away lives and children's educations.

Yoooouuuu pooooor idealistic dreamer ! That's your (our) goal but not "their" goal.

cdev · « **Reply #226 on:** December 13, 2018, 06:26:13 pm »

I shouldn't have complained about those examples, they are NOT particularly expensive. Also, TAPR certainly, being a nonprofit, deserves to make enough money back on their projects to pay for them.

So please let me apologize.

mark03 · « **Reply #227 on:** January 26, 2019, 03:37:19 am »

Quote

Code: [Select]
// $STATUS // 1: (Maybe 10 MHz bad, based on packrat docs) // 2: (Maybe PPS bad, based on packrat docs) // 3: Antenna is bad? 0=good // 4: Holdover duration (secs) // 5: Number of sats tracked (different than, but within 2 of $EXTSTATUS, perhaps only counts channels 0-7???, range is 0-8)

Quote from: cdev on August 15, 2017, 04:43:11 pm

Does it ever show a 1 in the third field (I think) of the $STATUS sentence? If not you are good to go as far as it being recognized.

I realize this is old now, but it might be useful for someone googling the thread. My Trueposition is connected through a 2-port GPS splitter (Instock PD5120). The TP is on the port with the DC block, while another GPS receiver is supplying DC bias to the antenna through the other, unblocked port. My $STATUS messages always have a 1 in the third field, but so far the board appears to be working fine. I tested with Lady Heather for a few days last fall, and have now reconnected it through the splitter with a QFH timing antenna in the attic. Currently doing a 24-hour survey just in case, although it was already locked.

cdev · « **Reply #228 on:** January 26, 2019, 03:52:16 am »

It sees that no current is being drawn so it registers 1 for that reason. In this case, telling you your DC block is working properly

Towger · « **Reply #229 on:** January 26, 2019, 06:47:48 am »

Yes. Some GPS splitters have a '200 Ohm Dummy Load' on each output to simulate an antenna, keeping the receiver happy. The downside is they also require external power.

cdev · « **Reply #230 on:** January 26, 2019, 05:03:39 pm »

A GPS distribution amp wouldn't need much gain, just very low noise, so could probably have a number of diodes to pull it from any of the GPSs that were on.

Quote from: Towger on January 26, 2019, 06:47:48 am

Yes. Some GPS splitters have a '200 Ohm Dummy Load' on each output to simulate an antenna, keeping the receiver happy. The downside is they also require external power.

mark03 · « **Reply #231 on:** February 06, 2019, 02:15:56 am »

Has anyone tried running their Trueposition board through gpsd, then pointing Lady Heather at gpsd instead of the serial port? I read somewhere that gpsd was read-only for clients, which I'm assuming would be a problem because LH needs to issue commands to the board. But I don't know that for sure. LH supports gpsd as one of its connection options, I just don't know if that will work for the Trueposition boards.

Reason for doing this is that I'd like to use the same Raspberry Pi as an NTP or PTP server, besides running the LH front-panel status display.

If the gpsd route doesn't work, it should still be possible to use the PPS signal as a reference clock, but use public NTP servers on the Internet to get "in the ballpark." Once the time is known +/- 0.2 seconds or so, NMEA is no longer required, only PPS. Still, it would be kind of nice for this to keep working during a network outage.

BTW, I'm running LH at 800x480 on this 4.3" LCD with the Adafruit adapter and it works great: https://www.buydisplay.com/default/4-3-800x480-ips-tft-lcd-module-all-viewing-optl-touchscreen-display
(The 5" display was too big for the front panel of the Extron distribution amp I'm building my GPSDO into.)

texaspyro · « **Reply #232 on:** February 06, 2019, 02:39:02 am »

Heather has the ability to echo the device data to a serial port as NMEA format messages. Check out the /ek command line option.

I doubt GPSD knows how to talk to the Truposition GPSDO, but have not checked.

mark03 · « **Reply #233 on:** February 06, 2019, 02:58:20 am »

Aha, even better! If I can abuse the /ek option to have Heather dump the NMEA messages into a named pipe instead of a serial device, I should be able to get gpsd to read from there.

It would be nice to get rid of gpsd altogether, but I was planning to use chrony for NTP and it looks like chrony needs gpsd via a Unix-domain socket, or a shared-memory segment.

mark03 · « **Reply #234 on:** February 10, 2019, 11:20:42 pm »

Does anyone know the optimum load impedance for the Trueposition's 10-MHz and PPS outputs? 50 ohms? I suppose the fact that they're brought out to coaxial connectors would imply this, but I wanted to make sure before connecting to my modified Extron distribution amp.

cdev · « **Reply #235 on:** February 14, 2019, 05:47:59 am »

That worked fine for me. I found I didn't even need to modify it at all. The TTL also works for 1pps.

I use el-cheapo 50 ohm terminators from ebay.

YMMV.

tkamiya · « **Reply #236 on:** February 14, 2019, 12:55:56 pm »

It was also supplied with a short piece of 50 ohm coax. That implies 50 ohm also.

cdev · « **Reply #237 on:** February 14, 2019, 03:00:57 pm »

GPS equipment is standardized on 50 ohm but since 75 ohm UHF-friendly coax (RG-6) is so much more widely available and cheaper, at least one manufacturer suggests its use in antenna installations. Since the typical GPS antenna has an LNA the coax doesn't matter so much, as long as its not really bad coax. One thing though, adding at least one ferrite bead at the antenna side sometimes improves the quality of the pattern in RTKlib markedly especially with patch antennas.

mark03 · « **Reply #238 on:** February 14, 2019, 08:40:43 pm »

I'm using 50-ohm input termination now and it works fine. I had removed the 75-ohm input resistors, set the project aside, then forgot I still needed to replace them. Without a load the 2x gain of the Extron amps caused clipping.

@cdev when you say "cheap 50-ohm terminators from Ebay" do you mean the unused Extron inputs? I'm feeding mine internal to the box, so no need for BNC terminations. If you meant the outputs... beware that a 50-ohm load on all three outputs of a single CLC409 may load it beyond the datasheet recommendation.

I had hoped to post pictures of my finished Extron GPSDO box this week but somehow in the shuffling around I managed to zap my Raspberry Pi's LCD panel. Now waiting on another from China

cdev · « **Reply #239 on:** February 19, 2019, 03:55:36 am »

To date I've only used a single channel and actually, I have put them on the unused Extron outputs of the channel I am using. I haven't looked at the waveform doing that, I guess it would be a problem because of the CW nature of a 10 MHz carrier? (as opposed to video signal) That totally makes sense, thank you will need to check.

I actually don't have any instruments yet that use the 10 MHz properly. I may use PicDiv and PicPet in a modular fashion to give me a few other frequencies, thats basically my plan right now, when I have the time to do it.

db5fp · « **Reply #240 on:** March 17, 2019, 02:57:12 pm »

Hi,

maybe I didi not understand the cammands of this device or I fill in the wrong commands.
I'm trying to save my location to FLASH to get a quick startup.
But after $UPDATE FLASH or dubble "EE" in Lady Haether the Position ist still $SURVEY 34733333 135350000 0 36 28800

Any hints?

Best regards

Squiddaddy · « **Reply #241 on:** March 18, 2019, 01:26:24 am »

Use $SETPOS with your correct coordinates. Or just run a $SURVEY for at least 8 hours.

db5fp · « **Reply #242 on:** March 18, 2019, 08:31:37 am »

Hi,

Thanks! Along survey and afterwarts $UPDATE FLASH solved the problem.

best regads
Frank

Dek · « **Reply #243 on:** March 28, 2019, 06:59:12 pm »

Hi,

I've just received my ebay source trueposition GPSDO and are trying to assimilate the various sources of info including ratpack info and this thread.
I can't decide if the TTL serial requires 3.3v to 5v level conversion, so will have to scope the data lines to find out but before I do that, what have others done?
Thanks,
Dek.

Towger · « **Reply #244 on:** March 28, 2019, 09:09:14 pm »

Quote from: Dek on March 28, 2019, 06:59:12 pm

I can't decide if the TTL serial requires 3.3v to 5v level conversion

https://www.eevblog.com/forum/index.php?topic=174531.msg2288070.msg#2288070

db5fp · « **Reply #245 on:** March 30, 2019, 09:04:48 am »

Sorry,

but still same problem. Position is not updated at startup.
It is still $SURVEY 34733333 135350000 0 36 28800

best regards

Frank

Towger · « **Reply #246 on:** March 30, 2019, 09:13:28 am »

When you start a survey it needs to finish before the position is updated. I think the default time is 8 hours.

CJay · « **Reply #247 on:** March 31, 2019, 06:25:11 pm »

And when the survey is complete, wow accurate.

It placed my cheap (very cheap) patch antenna to within a metre of the *indoor* windowsill it was sat on.

GNRFreak69 · « **Reply #248 on:** April 12, 2019, 01:17:01 am »

Ran across this thread a while back. Has anyone written a sketch for this other than the packrats hex file? Just don't wanna recreate the wheel.

-Jeff

rbm · « **Reply #249 on:** April 12, 2019, 01:27:27 am »

There are several:
https://www.eevblog.com/forum/metrology/lady-heather-v6-beta-for-windows-exe/?all
https://github.com/pigrew/trueposctrl
http://lists.febo.com/pipermail/time-nuts_lists.febo.com/2018-October/094480.html


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: ($40 GPSDO working well now, my mistake) (Read 61945 times)

Share me